In general, trials for improving operational efficiency are performed by evaluating the quality of driving a car or the like, to improve fuel consumption and to reduce accident rates in the transportation industry or the like. Driving evaluation methods which derive action data showing specific actions corresponding to sudden acceleration, sudden deceleration, sudden turn of a steering wheel, and the like, on the basis of driving data (such as speed data, acceleration data, and angular acceleration data) obtained from sensors and the like mounted on a vehicle are known as such driving evaluation methods. In addition, a technology for performing driving evaluation by using actual driving action data based on driving data and estimated action data based on road configuration data is known in such a type of driving evaluation technologies (Refer to following Patent Document 1 and 2).
On the other hand, a technology for performing driving evaluation by deriving smooth appropriate speed data denoting an ideal driving condition on the basis of actual speed data and comparing the appropriate speed data to actual speed data is proposed (e.g., refer to the following Patent Document 3). In this technology, a Fourier transformation is performed on the actual speed data to obtain a smooth curve composed of the sum of trigonometric functions as the above-described appropriate speed data, and driving evaluation is performed on the basis of a degree of divergence of the both data.